Four holes (1109A through 1109D) were cored at Site 1109, recovering a composite profile down to a depth of 802 mbsf. Physical properties measurements included MST readings on unsplit cores as well as index properties on discrete samples from split cores. Depending on the induration of the sediment, thermal conductivity was measured from either unconsolidated whole cores or discrete rock slices. Low recovery and fragmentation of core in the deepest part of the recovered succession precluded the use of the MST from Core 180-1109D-43R downward. Both undrained shear strength and unconfined compressive strength were measured in the uppermost 220 mbsf of the succession recovered from Holes 1109B and 1109C.
All index properties data are shown in Table T15. Gamma-ray attenuation (GRAPE) density measured on unsplit cores are found to agree well with the bulk sediment density determined from discrete sediment and rock samples from Holes 1109A through 1109C (Figure F75A), but not for the RCB-cored Hole 1109D. A full compilation of GRAPE data is presented with the MST measurement data in ASCII format on the accompanying LDEO CD-ROM.
Bulk density increases gradually from about 1.50 g·cm-3 near the seafloor to 1.75 g·cm-3 at ~120 mbsf (Fig. F75). Below this interval, density data decrease to values around 1.70 g·cm-3 at ~250 mbsf but are characterized by high scatter. This zone (~120-250 mbsf) corresponds to lithostratigraphic Units II and III (see "Lithostratigraphic Unit II" and "Lithostratigraphic Unit III"), and was deposited rapidly (i.e., at rates of ~400 m/m.y.; see "Biostratigraphy"). From ~250 to 374 mbsf, an almost linear increase in density exists from ~1.50 to 1.80 g·cm-3 (Fig. F75A). Below 380 mbsf, the GRAPE data from the RCB-cored Hole 1109D are consistently lower than the bulk density determinations from index properties measurements, although the observed trends of each remain similar. Lower densities from ~340 mbsf to 390 mbsf occur within the predominantly claystones and clayey siltstones of lithostratigraphic Unit V (see "Lithostratigraphic Unit V"). Density values increase with depth between ~390 and 670 mbsf, ranging from 1.70 to 1.90 g·cm-3 (Fig. F75). This interval is within lithostratigraphic Units VI and VII, which are characterized by relatively high sedimentation rates (231 m/m.y.; see "Biostratigraphy"). Below ~670 mbsf, the density data vary widely within the neritic, lacustrine, lagoonal, and subaerial deposits of lithostratigraphic Units VII to X. These units consist of siltstones and conglomerates. The GRAPE densities range from 1.50 g·cm-3 to values greater than 2.20 g·cm-3, the latter of which occur in conglomeratic units containing igneous rock clasts within a calcareous matrix. No GRAPE densities were obtained from the dolerite flows of lithostratigraphic Unit XI because of extensive core fracturing. Laboratory results on discrete samples, however, yielded bulk densities of 1.95-2.00 g·cm-3 for the sandstones and grainstones of lithostratigraphic Unit VII and 1.80-2.20 g·cm-3 for the underlying lagoonal sediments of lithostratigraphic Unit VIII. Dolerite bulk densities were found to vary between 2.50 and 3.00 g·cm-3 (Fig. F75A).
By comparison with bulk densities from the index properties measurements, GRAPE densities were generally found to be 0.2 to 0.3 g·cm-3 lower in cores recovered from Hole 1109D (i.e., beneath 380 mbsf in Fig. F75A). This discrepancy most likely relates to the decrease in diameter of the core produced by the RCB technique of drilling. However, correction of the GRAPE density data for the difference in both the core diameter (dXCB/dRCB = 1.15789) and the gamma-ray path length (factor 1.1726; following Boyce, 1976) failed to provide a satisfactory result. An empirical factor of 1.11691 led to a data shift resulting in excellent agreement between the GRAPE and index properties data. However, no theoretical justification can be found for this factor. The uncorrected GRAPE densities from the RCB, presented in Figure F75A, underestimate the bulk densities as determined by the index properties measurements. As a result, the GRAPE bulk sediment density should be used with caution.
Grain density was found to vary little with depth in the uppermost 380 mbsf, ranging from 2.55 g·cm-3 to 2.75 g·cm-3 (Fig. F75B). There is an overall but slight decrease in grain density in this interval, apparently independent of either lithology or grain size (Fig. F75). An increase in grain density with depth can be seen below 380 mbsf and may be associated with the higher abundance of volcaniclastic sands within lithostratigraphic Unit VI. Values between 2.65 and 2.75 g·cm-3 occur from 380 to 580 mbsf. Below this interval, grain density data are characterized by high scatter and an overall increase to values of up to 2.85 g·cm-3. Carbonate can only partly explain the overall grain density trend, being pertinent for the carbonate increase within the section from 420 to 580 mbsf (see "Organic Geochemistry") and with the carbonate-rich packstones and grainstones of lithostratigraphic Unit VII (580-670 mbsf; Fig. F75B). The change in the grain density trend at around 380 mbsf is likely related to a change in provenance, as suggested from clast compositions and paleomagnetic evidence (see "Lithostratigraphy" and "Paleomagnetism"). In the lowermost part of the recovered succession, wide variations in grain density reflect the lagoonal and subaerial deposits of Units IX through XI (Fig. F75B, "Lithostratigraphic Unit IX" and "Lithostratigraphic Unit X").
Porosity vs. depth strongly reflects bulk density variations, as expected from the method used for determination (see "Index Properties Measurements" in "Physical Properties" in the "Explanatory Notes" chapter). As illustrated in Figure F76, the uppermost part of the recovered section shows a strong reduction in pore space from an initial value of nearly 80% at the seafloor to values of 55%-60% at ~130 mbsf. The observed range in porosity corresponds approximately with what is known for the compaction of fine-grained marine sediments (e.g., Brückmann, 1989; Athy, 1930). Porosities are higher than expected in the intervals between ~130-280 mbsf and ~340-540 mbsf (Fig. F76). The first interval has porosities of 60%-68%, and the second has values between 55% and 62%. Similarly, results from anisotropy of magnetic susceptibility (AMS) indicate a smaller amount of uniaxial shortening parallel to the core axis in these intervals than in the overlying units (see "Magnetic Susceptibility"). High porosity zones, the AMS data, and the undrained shear and unconfined compressive strength are all consistent with the presence of underconsolidation zones within the recovered section. Underconsolidation and possible overpressuring of the anomalous intervals may be a consequence of rapid sedimentation rates (~225-312 m/m.y; see "Sedimentation Accumulation Rate") and overlying low-permeability sediments hindering fluid escape from these intervals (Fig. F76). Although no in situ pore pressure or permeability tests have been conducted, this hypothesis is supported by the existence of high-frequency fine-grained sediments at ~105-125 mbsf and 310-340 mbsf above the high-porosity intervals (see "Lithostratigraphy"). These clay-dominated sediments may well have prevented fluid expulsion from the underlying, rapidly deposited, coarser grained turbidites. The lowermost part of the porosity curve is controlled by the lithology, showing a distinct decrease down to 30%-40% within Units VII and VIII (packstones, siltstones, and claystones) but an increase of up to ~45% in the sandstones and conglomerates of Units IX and X (see "Lithostratigraphic Unit IX" and "Lithostratigraphic Unit X").
The P-wave velocity was measured using the P-wave logger (PWL) on the MST, the PWS1 and PWS2 insertion probe system, and the PWS3 contact probe system. The PWL logging on unsplit core yielded poor quality results; therefore, only the PWS data is presented here. All PWS data can be found in Table T16.
Transverse (i.e., across the core axis) and longitudinal (i.e., along the core axis) P-wave velocities were measured from Holes 1109A through 1109D. Below 270 mbsf (i.e., from Core 180-1109C-28X downward), the sediment was well lithified, requiring velocity measurements to be conducted on rock cubes in the x, y, and z directions. Data are presented in Figure F77.
Velocities obtained above 100 mbsf yielded values between ~1500 and 1650 m·s-1, as expected for poorly consolidated marine sediments. The velocities in the x and y (transverse) and z (longitudinal) directions show a linear increase with depth (Fig. F77). At ~560 mbsf, longitudinal velocities increase to ~1900-2050 m·s-1. Below 560 mbsf, a prominent increase in velocity is seen, with values ranging from >2000 m·s-1 to almost 3900 m·s-1 (shaded area in Fig. F77). These high velocities, particularly those in excess of 3000 m·s-1, correlate strongly with the existence of calcareous and siliceous cements (e.g., at 390 mbsf). In general, the steady increase in carbonate content with depth, from 420 to 480 mbsf (see "Inorganic Geochemistry"), corresponds to the general increase in velocities with depth. The abrupt increase and scatter in the velocities within the packstones and sandstones of lithostratigraphic Unit VII relates to their carbonate content (see "Lithostratigraphic Unit VII"). Velocities drop back to values around 2000 m·s-1 at a depth of ~680 mbsf and do not exceed 2300 m·s-1 within samples from the lagoonal siltstones and sandstones of lithostratigraphic Unit VIII. At the bottom of Hole 1109D, dolerite with velocities of 5000-6000 m·s-1 was measured.
Triaxial seismic velocity measurements indicate that the transverse and longitudinal velocities typically differ by <5% (Fig. F78). Nevertheless, there is a considerable amount of scatter within the data shown in Figure F78A and F78B. From 300 to 565 mbsf, the anisotropies are consistently skewed by ~5% toward higher transverse velocities. This interval correlates with the claystones and siltstones of lithostratigraphic Units IV, V, and VI. From ~580 to 700 mbsf, the anisotropy increases to values in excess of 15%, and is associated with the carbonate-rich sandstones, packstones, and grainstones of lithostratigraphic Unit VII and the claystones and siltstones of lithostratigraphic Unit VIII. The scatter within these latter units may be related to carbonate-rich and organic-rich intervals. It is unclear as to why the average transverse velocities should be greater than the longitudinal velocities; however, this effect is observed for claystones, siltstones, sandstones, and conglomerates alike.
Thermal conductivity on unsplit cores was obtained for all material recovered from Holes 1109A, 1109B, and 1109C. For RCB cores in Hole 1109D, measurements were carried out on discrete samples. Depending on core quality and time available, the number of measurements per core varied from two to six (Table T17). The reported value per measurement is an average of the repeat measurements. For the purpose of comparison between the two methods (and probes), thermal conductivity was determined on several cores both before and after splitting. The results were consistent and assured confidence in the data sets. This observation is supported indirectly, because no shift in the data is observed at the Hole 1109C-Hole 1109D transition at ~375 mbsf (Fig. F79).
Above 100 mbsf, thermal conductivities are 0.8-1.1 W·m-1·ºC (Fig. F79), increasing linearly with depth. From 100 mbsf to nearly 280 mbsf, thermal conductivity is widely scattered (0.7-1.1 W·m-1·ºC). This zone corresponds to Units III and IV and, more importantly, to the high porosity zones discussed above. Alternatively, some of the scatter may be related to using the needle probe in cores that are semilithified. In such cores, the coupling between sediment and probe is sometimes compromised (see "Thermal Conductivity" in "Physical Properties" in the "Explanatory Notes" chapter). Below 280 mbsf, thermal conductivities appear to follow a linear trend with depth. At ~300 mbsf, typical values are ~1.0 W·m-1·ºC, whereas at around 600 mbsf (i.e., the sandstones and packstones of lithostratigraphic Unit VI) an overall increase to ~1.35-1.5 W·m-1·ºC is observed. The lagoonal and subaerial deposits underlying Unit VII are thermally less conductive, ranging between 1.0 and 1.15 W·m-1·ºC (Fig. F79). This decrease in conductivity corresponds most likely to the inhomogeneous composition of the sediments that contain goethite nodules and clasts of various rock types. Thermal conductivities increase to values between 1.4 and 2.0 W·m-1·ºC for the dolerite recovered at the base of Hole 1109D.
Split cores from Holes 1109B and 1109C were used to determine un-drained shear strength and unconfined compressive strength using the motorized miniature vane-shear device and the pocket penetrometer, respectively. These strength parameters could not be measured below 220 mbsf because of increasing induration of the sediment. The data are presented in Figure F80, and listed in Table T18.
Undrained shear strength (Su) shows a slight linear increase with depth, rising from 10 kPa near the seafloor to ~30-40 kPa at 50 mbsf (Fig. F80). This depth range is characterized by the nannofossil-rich calcareous sands of lithostratigraphic Subunit IA. Within the clay-dominated turbidites of lithostratigraphic Subunit IB, the strength increases more rapidly to values up to 125 kPa at ~72 mbsf. From the top of Unit II (83.4 mbsf) downward, strength decreases to values as small as 10 kPa, which can be partially explained by the effects of bioturbation, volcaniclastic enrichment of the sediment, or drilling disturbance of the core.
If the unconfined compressive strength (2 Su) is compared to the undrained shear results, similar trends can be recognized. A linear increase in strength is observed over the uppermost interval between ~20 and 120 mbsf, with strengths ranging from 30 kPa to more than 200 kPa. Below ~120 mbsf, even the unconfined compressive strength drops to values close to those obtained near the seafloor (Fig. F80). With the exception of the outliers, in the interval below 130 mbsf strength is generally ~40-50 kPa and does not exceed 100 kPa.
From a least-squares linear fit to the strength data, the extrapolated downhole trend was found to vary between 0.30 and 2.57 kPa·m-1 (from undrained shear strength, with an average of 1.18 kPa·m-1) and 1.60 kPa·m-1 (based on the pocket penetrometer data). The state of sediment consolidation can be estimated by comparing the range of strengths measured relative to a normally consolidated sediment, expressed as the ratio of undrained strength (Su) to effective overburden stress (P0'). The effective overburden stress is calculated using grain and pore-water density and the thickness of the overlying sedimentary succession, if hydrostatic pore-fluid pressures are assumed. Normally consolidated sediments show Su/P0' ratios of 0.2 (e.g., Mesri, 1975), as shown by the dashed red line in Figure F80. Note that the dashed red line would have a steeper gradient if the pore-fluid pressure were above hydrostatic. However, it can be inferred that the upper part of the sediments, namely lithostratigraphic Units I and II, have undergone normal consolidation. On the other hand, the sediment beneath ~130 mbsf is underconsolidated, as has been suggested from the porosity and bulk density (see discussion in "Density and Porosity").
Magnetic susceptibility reflects changes in magnetic mineralogy (e.g., lithologic variations) and was obtained routinely as part of the MST measurement of sediment cores from Holes 1109A through 1109D. The quality of magnetic susceptibility data is a function of reduced core diameter and/or increased disturbance to the core by the drilling process; therefore, it commonly degrades from APC- to XCB- to RCB-cored sections. The full magnetic susceptibility data set can be found as part of the MST compilation in ASCII format on the accompanying LDEO CD-ROM.
The magnetic susceptibility of the recovered section is displayed in Figure F81 and is compared with grain-size distribution and the remanent magnetic intensity. Because of this compressed scale, only the general trends can be identified. In particular, there exists a first-order difference between the high-amplitude susceptibility within the upper part of the section (80-380 mbsf) and the low-amplitude susceptibility within the middle part of the section (380-540 mbsf). The transition occurs within lithostratigraphic Unit V and marks the first input of metamorphic detritus into the basin. The high-amplitude magnetic susceptibility zone between 80 and 360 mbsf correlates with the presence of high-frequency, fine-grained turbidites. In contrast, the low-amplitude magnetic susceptibility zone (380-540 mbsf) relates to high-frequency turbidites that are characterized by abundant fine to coarse sands. Lithologically, these differing clay- and sand-dominant turbidite units have been termed distal and proximal, respectively (see "Lithostratigraphy"). Second-order trends are represented by the relatively high amplitude susceptibility variations occurring at the base of lithostratigraphic Unit VI and the transition between lithostratigraphic Units VI and VII (540-600 mbsf), the lagoonal and subaerial deposits on lithostratigraphic Units IX and X, and the dolerite at the base of Hole 1109D (Unit XI; 720-796 mbsf).
At the top of the succession, frequent ash and sand layers interbedded with nannofossil clays correspond with spikes on an otherwise uniform curve. Highly variable but generally increased susceptibilities were obtained from 80 and 170 mbsf, which strongly reflects the increase in volcaniclastic input to the section (i.e., lithostratigraphic Unit II) and possibly the dominance of silts and clays (Fig. F81A). Within Units III, IV, and V, no obvious correlation between grain size and susceptibility can be found (see "Site 1109 Core Descriptions"). Apparently, some, but not all, of the spikes in the susceptibility curve correlate with the location of coarse-grained sediment. On the other hand, the presence of two high-frequency turbidite intervals (in Units II and V; see "Lithostratigraphic Unit II" and "Lithostratigraphic Unit V") correspond to relatively high susceptibilities. However, data scatter is more likely related to different sources of material within these sediments, namely metamorphic components and mixed-layer clays (see "Lithostratigraphy"). Sediments above Unit V consist of clays and silts including components of altered calc-alkaline volcaniclastic material and metamorphic detritus, with smectite and mixed-layer clays. Below ~370 mbsf, magnetic susceptibility is rather uniform and near zero in lithostratigraphic Unit VI. Here, turbiditic sands and muds derived from a dominantly fresh basalt-andesite volcanic source with minor neritic carbonates show little or no alteration, suggesting a major change in provenance.
Unit V is a transitional interval (350-380 mbsf), with both variable grain size and scatter in magnetic susceptibility. The observed relationships imply that the magnetic material derived from the clastic source regions of both the lower uniform and the upper, more susceptible interval appears both in fine- and coarse-grained deposits. The remarkable change at ~370 mbsf is reflected in the results from both the natural gamma ray (NGR) logging data as total gamma ray count (HSGR) (refer to Fig. F88) and the remanent magnetization intensity (see both Fig. F81 and "Paleomagnetism"). The packstones of lithostratigraphic Unit VII (~560-645 mbsf) show consistently higher magnetic susceptibility compared with the overlying lithostratigraphic units, whereas the underlying claystones are uniformly close to zero. Both the altered siltstones of proposed lagoonal origin (including Fe-rich concretions scattered within; see "Lithostratigraphic Unit VII") and the conglomerates with variable clast lithologies exhibit higher susceptibilities, often exceeding 0.015 SI.
By comparing the magnetic susceptibility with the remanent magnetization intensity (see "Remanent Magnetization,"), it is possible to determine if the mineralogy carrying the magnetic remanence is also responsible for the susceptibility. Figure F81 compares the magnetic susceptibility, grain size, and remanent intensity. Within the sequences characterized by a metamorphic clastic component (0-380 mbsf), there exists a direct correlation among the susceptibility, NGR (Fig. F82; also see "Natural Gamma Ray"), and remanent intensity. We conclude that, in general, the mineralogy controlling the magnetic susceptibility is the same as that controlling the remanent magnetic intensity. Further, the correlation between NGR variations and the magnetic susceptibility suggests that the clays and sands containing radioactive material also are rich in ferromagnesian minerals.
The NGR count was recorded on the MST, and a composite log is shown in Figure F82. The full data set can be found as part of the MST compilation in ASCII format on the accompanying LDEO CD-ROM. Superposed on the NGR count is the HSGR logging data. The excellent agreement between the MST gamma-ray data and the logging gamma-ray data supports use of the logged data in regions of poor core recovery. Although there is significant scatter, a number of general trends can be recognized in the data. From the seafloor to 380 mbsf, the NGR count is remarkably similar in form to the magnetic susceptibility (Fig. F82) and the bulk density (Fig. F75). This same range consists of the high-frequency distal turbidites referred to above. From 380 to 580 mbsf, the NGR count is rather uniform although the absolute level is ~20 counts/s. The local NGR count maxima centered on 140 and 320 mbsf have about the same value. Further, both the majority of the carbonate-rich lithostratigraphic Unit VII and the lagoonal/swamp deposits of lithostratigraphic Unit VIII are associated with high NGR counts.
In general, there is no straightforward relationship between high and low NGR with the presence of clay/silts and sands, respectively. The reason why this simple distinction cannot be made at Site 1109 may be that either the sequences are poorly sorted or that the lithology varies on a rather small scale. The characteristic highs at 80-180 mbsf and 260-370 mbsf (see Fig. F82) do not correspond to obvious lithostratigraphic changes or unit boundaries (Fig. F82). The usual high/low NGR relationship is confounded at Site 1109 by the presence and inter-mixing of radioactive sands and clays. For example, a ~12-m-thick sand layer interpreted at 219-231 mbsf is marked by a strong NGR peak. This sand layer was not recovered except for minor amounts in the core catcher (for details, see "Lithologic Analysis"). In contrast, the overlying clay unit (210-219 mbsf) of similar thickness as the sand is characterized by a low NGR count. The abrupt change from variable amounts of radioactive sediment components in the upper part of the succession to a uniformly lower level at ~370 mbsf may correspond to the change in provenance mentioned above (see "Lithostratigraphy").
An important application of MST measurements is to determine the correlation between multiple holes at this site (see detailed section on "Composite Depths"). For this purpose, GRAPE density, magnetic susceptibility, and NGR were used to correlate between Holes 1109A, 1109B, and 1109C, and magnetic susceptibility to correlate between Holes 1109C and 1109D. The correlation is based on recognizing characteristic features in the data sets. For example, Figure F83 shows a small, relatively high density, calcareous sand unit at 2.75-3.25 mbsf in the Hole 1109A core. A similar density anomaly exists within the core at Holes 1109B and 1109C, suggesting (1) minimal offset between Holes 1109A and 1109B and (2) a 0.4- to 0.5-m offset between Holes 1109B and 1109C (shaded zone, Fig. F83). Similarly, the magnetic susceptibility contains distinct peaks at 5.2-5.5 mbsf and 9-10 mbsf, and the NGR has a sequence of characteristic peaks at 7.45-10 mbsf and at 10.8-11.5 mbsf (shaded zones; Figs. F84, F85). Unlike the GRAPE density interhole correlation, these data suggest a minimal offset between Holes 1109A and 1109C.
The magnetic susceptibility data were also used to correlate between Holes 1109C and 1109D (Fig. F86). A characteristic arrangement of peaks exist between 355 and 356 mbsf in Hole 1109C cores. This same pattern exists in Hole 1109D but between 355.7 and 356.5 mbsf, suggesting an offset of 0.5 m (shaded zone, Fig. F86).
Physical properties measurements obtained from cores and downhole logging data from Site 1109 provide complementary data sets. Logging data yield denser measurements and fill gaps in areas of low core recovery. Core measurements provide data in areas where logging is effectively precluded by either the drill pipe (above ~120 mbsf) or washouts. It is interesting to note that the volume of rock involved in a downhole logging measurement is on the order of 30 dm3 (1 cubic foot), whereas that of a physical properties measurement is on the order of 1 cm3. In addition, physical properties measurements can be directly related to other shipboard core descriptions and measurements. Index properties determinations of bulk density are in good agreement with those from the logging run, although logging data show slightly higher magnitudes (Fig. F87). Significantly, logging data confirm a high-density zone around 219-233 mbsf that was poorly defined by physical properties measurements because of low recovery (see "Log Unit L2"). Logging data below 680 mbsf indicate very low bulk densities, but data quality in this zone may have been degraded by washouts.
Core physical properties measurements of longitudinal (z-direction) velocity show excellent agreement with data from downhole sonic velocity logs (Fig. F86). Logging data reveal a high-velocity interval between ~220 and ~233 mbsf, coincident with the high-density zone and poor recovery.